Environmental Discretionary Reward Scheme SP Energy ... › userfiles › file › ...Environmental Discretionary Reward Scheme 2 SP Energy Networks 2015–2023 Business Plan Updated

Environmental Discretionary Reward Scheme 2

SP Energy Networks 2015–2023 Business Plan Updated March 2014

Annex Assessment of Overhead Line Performance During Severe Storms EA Technology April 2012

CONFIDENTIAL REPORT

Prepared for:

Scottish Power Energy Networks

Assessment of Overhead Line Performance during severe storms

Author :

Andrew J McHarr ie

Repor t No : 82880

CONFIDENTIAL - This document may not be disclosed to any person other than the addressee or any duly authorised person within the addressee's company or organisation and may only be disclosed so far as is strictly necessary for the proper purposes of the addressee which may be limited by contract. Any person to whom the document or any part of it is disclosed must comply with this notice. A failure to comply with it may result in loss or damage to EATL or to others with whom it may have contracted and the addressee will be held fully liable therefor.

Care has been taken in the preparation of this Report, but all advice, analysis, calculations, information, forecasts and recommendations are supplied for the assistance of the relevant client and are not to be relied on as authoritative or as in substitution for the exercise of judgement by that client or any other reader. EA Technology Ltd. nor any of its personnel engaged in the preparation of this Report shall have any liability whatsoever for any direct or consequential loss arising from use of this Report or its contents and give no warranty or representation (express or implied) as to the quality or fitness for the purpose of any process, material, product or system referred to in the report. All rights reserved. No part of this publication may be reproduced or transmitted in any form or by any means electronic, mechanical, photocopied, recorded or otherwise, or stored in any retrieval system of any nature without the written permission of the copyright holder. © EA Technology Ltd April 2012

EA Technology Limited, Capenhurst Technology Park, Capenhurst, Chester, CH1 6ES; Tel: 0151 339 4181 Fax: 0151 347 2404

http://www.eatechnology.com Registered in England number 2566313

Project No: 82880

Assessment of Overhead Line Performance during severe storms

EA Technology Assessment of Overhead Line Performance during severe storms

Project No. 82880

1

Summary This document is an asset performance review of Scottish Power’s Scottish network during

the storms of December 2011 and January 2012. It evaluates the four storms of this period

and compares them against the storm event of the 26th December 1998. The report

concludes that the storm event of the 3rd January 2012 was the most similar to the storm

event of 26th December 1998.

The comparison of the two storms allows the Scottish network’s performance to be

evaluated since the introduction of Scottish Power Energy Network’s (SPEN) improvement

policies post 1998. The Storm event of 2012 saw a total reduction of all faults of 87% over

the 1998 event.

Whilst the storm event of 2012 is deemed to be comparable with the storm event of 1998 the

main differences were:

Dumfries had proportionally less faults in 2012 and central and Fife had

proportionally more when compared to 1998. This observation could suggest that

the 2012 storm’s trajectory was slightly higher than that of 1998.

The storm event of 2012 made a quicker impact on the network than 1998; however

the maximum number of customers without supply was reduced by 18% compared to

1998. The 2012 storm clear up phase was also 39% quicker than 1998.

The CML result for the storm event of 2012 was 64% lower than 1998, and the CI

result was 48% lower.

The major cause of damage during the 1998 storm event was attributed to trees at 85%.

During the storm event of 2012 only 61% of the damage faults were attributed to trees. It is

known that a number of trees fell during the 2012 storm that had previously been assessed

as healthy. The failure of these trees is suspected to be primarily due to the combination of

excessive winds and saturated ground due to abnormally high rainfall (Figure 19).

The upgrading of OHLs and the carrying out of tree clearance to ENA TS 43-8 and ENA

ETR 132 has had an impact on the amount of faults over the duration of a comparable storm

event. The upgrading of OHLs and the frequency of tree clearance should be continued as

per SPEN’s current strategy, with the exception of spur lines with large numbers of

connected customers, which should also be considered for inclusion into the improvement

program.

It was found that the L10 Overhead Line (OHL) construction performed well during the 2012

storm event; therefore an assessment should be carried out into whether the L10

construction could be utilised within the current severe weather areas, thus allowing more

OHL circuits to be rebuilt at a lower cost.


Project No. 82880

2

Further work should also be undertaken to refine the severe weather map for SPEN’s

Scottish network. The STP21 project S2641 is undertaking work to develop new wind/ice

load maps, which are expected to be less onerous than those in ENA TS 43-40. The

refinement of the severe weather map could also more accurately target the correct

construction for the environment.

Consideration should be given into an insulated conductor construction, as a targeted

solution for poorly performing circuits; or ones that are experiencing difficulties in obtaining

permissions for a re-build and/or ETR 132 compliance.

No trees fell on any of SPEN’s re-built OHLs, which are constructed to ETR 132. This would

suggest that ETR 132 should be considered on a wider basis than re-build circuits. For

circuits that will struggle to achieve ETR 132 compliance, consideration should also be given

into declaring a percentage of the line as ETR 132 compliant; or dividing the circuit with ABS

switches either side of the anomoly and declaring either side as ETR 132 compliant and the

middle as non-compliant.

Pole condition should be added to the AIS inspections, as this will reduce the amount of

poles that need re-testing plus it will give a better understanding of the Health of the pole

asset class. As Ofgem move into the next price review (RIIO ED1) greater emphasis will be

placed on a risk-removed output measure for asset classes. DNOs will be required to take a

holistic risk-based approach to asset management, by reporting the risk of an asset class at

the beginning of the ED1 price review and, with investment, its risk will be at the end of ED1.

The important outcome at the end of the RIIO ED1 price review will be the Health and

subsequent risk of an asset class rather than how many kilometers of line were re-built. Risk

management processes like EA Technology’s CBRM (Condition Based Risk Management)

process can evaluate the risk of an asset class through age, environment, probability of

failure and criticality, however it is much more accurate if condition data is also supplied.

Major investment is required in the upgrading of LV open wire constructions to ABC, the LV

network was identified in the Boxing Day storm review report [1] as being in need of

refurbishment after the storm event of 1998. The investment will help reduce the amount of

LV faults during storm events plus it will drastically reduce the clear up duration and non-

planned CMLs, as the majority of the time identified on the customer interuption graph, was

due to LV circuits with low numbers of connected customers. Consideration should also be

given to renewing the first LV poles from OHL substations and replacing the first span with

ABC conductor during HV refurbishment. This should reduce the number of LV faults that

escalate into HV faults, plus it will reduce the need for subsequent HV shutdowns during LV

refurbishment, which in turn will help reduce ‘Planned CIs’.

1 SPEN is a member company of STP2 and as such will receive a copy of the findings from the S2641

project.


Project No. 82880

3

Contents Page

1 Introduction 4

2 Legal Obligations 6

2.1 ESQCR 6 2.1.1 ESQCR 2002 states in Part I, Paragraph 3 - (1) 6 2.1.2 ESQCR 2002 states in Part I, Paragraph 5: 6 2.1.3 ESQCR 2006 states in Part VA, Paragraph 20A: 7 2.1.4 ESQCR 2002 states in Part VII, Paragraph 23(1): 7

2.2 DPCR 5 7 2.2.1 SLC 45 7 2.2.2 SLC 44A 7

3 Major Storm Events 8

3.1 Boxing Day 1998 8 3.2 2011/12 Storm Events 8 3.3 Storm Event of 2006 11

4 Network Performance 13

4.1 HV Network Performance 13 4.1.1 Network Failures 15

4.2 LV Network Performance 16 4.2.1 Network Failures 17 4.2.2 LGC Programme 17

4.3 ABC 18 4.4 Ofgem Metrics 19

5 Improvement Policies 20

5.1 OHL Policies 20 5.2 Vegetation Management 22 5.3 Severe Weather Map 24

6 OHL Protection Policy 27

7 Poles 29

7.1 Installation 29 7.2 Condition 30

8 Targeted Solutions 32

8.1 BLX 32 8.2 Ericsson 33 8.3 Spacer Cable Systems 34

9 OHL Maintenance 35

9.1 Prioritisation 35 9.1.1 HV 35 9.1.2 LV 36

10 Conclusions 37

11 Recommendations 39

12 References 41


Project No. 82880

4

1 Introduction

EA Technology was requested by Scottish Power Energy Networks (SPEN) to undertake an

assessment of the performance of their Overhead Line Network in Scotland, during the

severe weather storms of late December 2011 and early January 2012. The objective of this

work is to determine if the Overhead Line Network investment undertaken since year 2000

has improved its reliability and resilience to severe storms.

To facilitate the objective the scope of the project was to:

1. Contrast and compare the OHL asset performance with previous storm events;

2. Identify potential asset performance improvements;

3. Make recommendations for performance Improvement.

SPEN’s Overhead line network experienced a major storm event on the 26th December

1998 and as a result of the subsequent panel of inquiry, a number of policies were

introduced that were designed to increase the network’s resilience. The panel of enquiry’s

report [1] concluded that the major cause of storm related damage during the storm event

was related to trees either falling or interfering with overhead lines.

The panel of inquiry assessed what and where Scottish Power could have done better, and

proposed that lessons learnt, would improve how Scottish Power work and what they will do

if a similar situation arises again.

The main areas of improvement were split into four categories:

1. Investment in the call centre to increase the number of calls it can handle.

2. Improve the processes and systems that are used to monitor and control the network

and to restore supplies.

3. A more rigorous policy for making sure there are no trees near strategic power lines.

4. Strengthen the network in appropriate areas.

This document sets out to determine the effect on the network during a major storm event, of

categories 3 and 4 since the storm event of 1998. A separate evaluation is currently being

carried out to assess the operational performance during the recent storms; therefore

category 1 and 2 will be covered in this report.

The ‘Rural Care’ policy was introduced in 1999, which set out to clear vegetation below and

within falling distance of strategic or poorly performing overhead line circuits. This policy

was updated in 2002 by introducing low and high risk zones. The low risk zone required a 2

metre clearance to be maintained from low risk vegetation, whilst the high risk zone required

clearance of falling distance for high risk trees.

March 2006 saw the introduction of a risk based vegetation management methodology

ETR132 [2] by the Energy Networks Association (ENA), SPEN currently strives to achieve

compliance with ETR132 on all of its overhead line rebuild projects.


Project No. 82880

5

Figure 1 – Example of a Resilient Clearance (Source: ETR132 [2])

SPEN also implemented OHL-01-008 [3] in 2002; this set out to improve network resilience

by introducing a number of overhead line standards that were more robust and therefore

more resilient than older standards such as BS1320 [4]. A new heavy construction standard

(L27) was created for HV overhead lines that were deemed to be within a severe weather

area.

For clarity, the networks described within this document are as per the definition within ENA

ER 43-3 [5]. An LV system is one which operates at a nominal voltage up to a 1,000 V; an

HV system is one which operates at a nominal voltage in excess of 1000 V but less than 22

kV; an EHV system is one which operates at a nominal voltage equal to or greater than 22

kV, but less than 132kV.


Project No. 82880

6

2 Legal Obligations

SPEN are bound by many statutory and regulatory requirements, the following have been

identified as being the most pertinent in regards to SPEN’s network resilience under storm

conditions.

2.1 ESQCR

The Electricity Safety Quality and Continuity Regulations 2002 (ESQCR) came into force on

31st January 2003 [6] and was amended in 2006 [7] and 2009.

2.1.1 ESQCR 2002 states in Part I, Paragraph 3 - (1)

“Generators, distributors and meter operators shall ensure that their equipment is:

(a) sufficient for the purposes for, and the circumstances in which it is used; and

(b) so constructed, installed, protected (both electrically and mechanically), used and maintained as to prevent danger, interference with or interruption of supply, so far as is reasonably practicable”.

The introduction of the OHL-01-008 [3] after the Boxing Day storm of 1998 introduced

severe weather constructions to increase network resilience. SPEN in conjunction with the

Met office have Identified areas that would be prone to severe weather (see 5.3 Severe

Weather Map). The identification of these areas allows SPEN to target, the correct OHL

construction for the local conditions and environment, as well as targeting its more costly

severe weather constructions more effectively.

2.1.2 ESQCR 2002 states in Part I, Paragraph 5:

“A generator or distributor shall, so far as is reasonably practicable, inspect his network with

sufficient frequency so that he is aware of what action he needs to take so as to ensure

compliance with these Regulations and, in the case of his substations and overhead lines,

shall maintain for a period of not less than 10 years a record of such an inspection including

recommendations arising therefrom”.

The Regulations give no guidance as to the methods or frequency of inspections; however

SPEN carry out AIS inspections on its LV and HV networks on a 6 year basis for poles

categorised as normal risk. Poles that are deemed to have a raised environmental risk shall

be inspected on a more frequent basis of 3 years. SPEN also carries out vegetation

inspection every 3 years which would also identify any major defects.

SPEN also carries out aerial inspections on an annual basis for its EHV network.


Project No. 82880

7

2.1.3 ESQCR 2006 states in Part VA, Paragraph 20A:

“A generator or distributor shall, so far as is reasonably practicable, ensure that there is no

interference with or interruption of supply caused by an insufficient clearance between any of

his overhead lines and a tree or other vegetation.”

SPEN carries out vegetation clearance to ENA TS 43-8 [8] on a 3 year cycle, every fourth

cycle (12 years) all OHLs will undergo tree clearance to ETR 132 [2].

2.1.4 ESQCR 2002 states in Part VII, Paragraph 23(1):

“A distributor shall ensure that his network shall be—

(a) so arranged; and

(b) so provided, where necessary, with fuses or automatic switching devices,

appropriately located and set,

as to restrict, so far as is reasonably practicable, the number of consumers affected by any

fault in his network.”

SPEN achieve regulation 23(1) by its implementation of its Overhead Line Protection Policy (OHPP)

that was introduced in 1993/94 (6 OHL Protection Policy).

2.2 DPCR 5

2.2.1 SLC 45

Ofgem’s Regulatory Instruction and Guidance (RIG) SLC 45 ‘Incentive scheme for quality of service’

places obligations on DNOs for reporting on quality of service. SPEN records and archives all of its

planned and unplanned outages in it’s PROSPER outage reporting system, which is then uploaded

into the National Fault and Interruption Reporting Scheme (NaFIRS). At present SPEN reports the

majority of its short interruptions data in bulk (as defined in ENA ER G43-3 [5]) due to a majority of

it automatic switchgear not possessing any signalling connections to SPEN control centres.

2.2.2 SLC 44A

Ofgem’s RIG SLC 44A ‘Network outputs regime’ places obligations for reporting on network outputs

such as CMLs, CI’s and Health Indexes. SPEN complies with SLC 44A through it’s PROSPER and

AIS systems.


Project No. 82880

8

3 Major Storm Events

3.1 Boxing Day 1998

Between the 26th December 1998 and the 4th January 1999 Scottish Power experienced the

worst storm event it had encountered in 30 years. The Boxing Day storm, as it became

known, was second only to the 1968 ‘Glasgow hurricane’ and covered a large area of

southern Scotland.

The storm event resulted in 212,900 customer interruptions and 7,622 faults that were

directly related to the storm event. Table 1 illustrates that the total CML score for the 1998

storm event was 94.11 and that the HV network was the largest contributor at 82.04.

Table 1 – CML and CI results for 26/12/1998 storm event

Figure 2 displays the main characteristics of the 1998 storm event upon which the initial

comparisons will be made to the storm events of 2011/12.

Figure 2 – Storm event of 26

th December 1998 (regions are an indicative representation)

3.2 2011/12 Storm Events

SPEN’s Scottish network was subjected to four significant storm events over 27 days from

the 8th December 2011 to the 3rd January 2012. The four storm events of 2011/2012 were

compared to the Boxing Day storm of 1998 to determine the storm that most closely

matched to form the basis of the evaluation. The four storm events of 2011/12 were 8th

December 2011; 13th December 2011; 28th December 2011 and 3rd January 2012.

CML due to the Storm Event CI due to the Storm Event

Voltage West Scotland East Scotland EN North West Scotland East Scotland EN North

EHV 8.61 0.1 8.71 1.88 0.04 1.92

HV 59.01 23.02 82.04 5.82 3.34 9.15

LV 2.5 0.86 3.36 0.22 0.07 0.29

Total 70.12 23.99 94.11 7.91 3.45 11.36

26th December 1998

Predominantly South West to Westerly winds.

Maximum recorded gust speed 103mph.

17 hour duration at mean wind speed above 30mph.

6 hour duration at mean wind speed above 50mph.

No Snow or Ice accretion.


Project No. 82880

9

Figure 3 – Storm event of 8th

and 13th

December 2011 (regions are an indicative representation)

Figure 4 – Storm event of 28th

Dec 2011 & 3rd

Jan 2012 (regions are an indicative representation)

Figure 4 shows that the storm event of the 3rd January 2012 is comparable to the

Boxing Day storm event in magnitude and trajectory. The two storm events were

synoptically similar as they both evolved from a rapidly deepening depression to the

west of Ireland, which then swept across Highland Scotland on an East-Northeast

trajectory [9].

Figures 2 to 4 also display the HV faults that occurred within each of SPEN’s Scottish

regions, which are Ayrshire & Clyde; Borders; Central & Fife; Dumfries; Edinburgh &

Lothians; Glasgow & Clyde North and Lanarkshire. Figure 6 illustrates these faults

for comparison, and identifies that the proportion of faults for the storm events of

08/12/2011 and 03/01/2012 are very similar to each other, plus they closely resemble

the storm event of 26/12/1998.

8th

December 2011

Westerly winds.



13th

December 2011

Westerly winds.



28th

December 2011

Initially South Westerly winds,

slowly veering Westerly then West

NorthWest



3rd

January 2012

Predominantly South West to Westerly winds.


6 hour duration at mean

wind speed above 50mph.



Project No. 82880

10

Figure 5 – 3

rd January 2012 storm trajectory

The main difference between the pie chart of the 26/12/1998 and the pie charts of the

08/12/2011 and 03/01/2012 is that Dumfries suffered proportionally more faults in 1998

whilst in 2011/12 Central and Fife had more faults. This observation could suggest

that the storm events of the 08/12/2011 and 03/01/2012 were similar in their trajectory

to the Boxing Day storm however the Boxing Day storm’s path travelled slightly lower

across SPEN’s Scottish Network.

Figure 6 – Fault Locations


Project No. 82880

11

Table 2 [10] illustrates that the storm gust speeds of 3rd January 2012 are the highest since

1998 at a majority of the weather stations, it can also be seen that the max gust speeds are

similar in magnitude.

Table 2 - Highest gust speeds 3 January 2012 (excluding mountain stations)

Station Elevation

(m.a.s.l.)

Max gust

speed (knots)

Highest gust since %

Edinburgh,

Blackford Hill 134 89 26 December 1998 93 knots 96%

Salsburgh,

Lanarkshire

277 84 26 December 1998 95 knots 88%

Islay, Port Ellen 17 84 Highest gust on record – previously 26 December

1998 83 knots

101%

Aberdaron, 95 81 8 January 2005 89 knots N/A

Drumalbin,

Lanarkshire

245 80 26 December 1998 82 knots 97%

Machrihanish,

Argyll

10 80 26 December 1998 82 knots 97%

Glasgow,

Bishopton

59 79 Highest gust on record – previously 28 January

2002 67 knots. Records from 1999 only

118%

Capel Curig,

Gwynedd

216 75 7 February 2011 75 knots N/A

Dunstaffnage, Argyll

3 75 8 December 2011 76 knots 99%

Edinburgh, Gogarbank

57 74 Highest gust on record – previously 28 January 2002 71 knots. Records from 1999 only

104%

Isle of Portland, Dorset

52 74 25 December 1999 80 knots N/A

Of the four storm events between the 8th December 2011 and 3rd January 2012

considered in this document, the storm event of the 3rd January 2012 will be the one

compared to the storm event of the 26th December 1998

3.3 Storm Event of 2006

On the 31st December 2006 SPEN’s Scottish network experienced a severe storm event that

resulted in widespread damage to the overhead network. SPEN requested KEMA review its

HV overhead line assets in terms of storm resilience [11]. KEMA searched Scottish Power’s

PROSPER reporting system to select the days between December 1998 and January 2007

that had experienced 50 or more storm related faults. This resulted in four storm events

identified in Figure 7.

Figure 7 – Summary of storm characteristics


Project No. 82880

12

The KEMA report is based exclusively on the performance of HV assets; therefore for

evaluation against the 2011/12 storm events, only HV fault data was used. Figure 7 displays

the storm event characteristics evaluated by KEMA and, as they are less of a match to the

Boxing Day storm than the storm event of the 3rd January 2012, none of these will be used in

the evaluation. Figure 8 illustrates the number of HV faults reported compared to the

maximum wind speed recorded during the storm events.

Figure 8 – HV Faults Vs. Maximum Wind Speeds

In conclusion, from the eight storm events illustrated in Figure 8 the 1998 and 2012

storm events are the most comparable.


Project No. 82880

13

4 Network Performance

The storm event fault data detailed in Table 3 was sourced from ‘Boxing Day Storm Divisional Report.pdf’ [1] and ‘Energy Networks North – Exceptional Conditions.pdf’ (Appendix 1). It details that the whole network suffered from 87% less faults than the 1998 storm event.

Table 3 – Storm Event Fault Data

Weather related Faults 26/12/1998 03/01/2012 % Reduction

LV 6,051 518 92%

HV 1,421 501 65%

EHV 150 17 89%

Total 7,622 1,064 87%

Transmission 179 28 84%

Table 3 illustrates that the HV and LV networks are, as would be expected, the major cause

of faults during both of the storm events. These networks are inspected on a less frequent

basis than the EHV networks; their inspection frequency is dependent on their environment

risk. HV and LV assets defined as Low risk are inspected on a 6 yearly basis whilst assets

defined as High risk are inspected every 3 years.

4.1 HV Network Performance

All statistical analysis of the HV network was carried out on information detailed in ‘HV OHL

Damage faults only 261298.xlsx’ and ‘HV OHL Damage faults only 030112.xlsx’, which

contained data on damage faults only. The data within the ‘HV OHL Damage faults only

261298.xlsx’ spreadsheet was based on data from the PROSPER corporate system. This

data is known to have been massively under reported due to a number of failures in the

TroubleCall and ICOND systems during the 1998 storm event. These figures have been

used for proportional evaluation, however it has to be recognised that when compared to the

‘Boxing Day Storm Divisional Report.pdf’ [1]; these are incorrect to a factor of 3 for EHV, 2

for HV and 8 for LV.

The storm event of the 26th Dec 1998 (Figure 9) had 44% mainline faults and 27% spur

lines. The storm event of 2012 in comparison with 1998 experienced a 31% reduction in HV

(Damage) faults, these were categorised as 56% mainline and 41% spur lines. The main line

and spur line figures are not comparable due to the large number of overhead lines

categorised as ‘Other’ in 1998. What it does display, is that the reporting of the circuits line

type has drastically improved with only 3% reported as ‘Other’ in 2012.


Project No. 82880

14

Figure 9 – Overhead Line types

Figure 10 illustrates the HV damage fault counts for the storm events of the 26 th Decemeber

1998, 8th Decemebr 2011 and the 03rd January 2012. It can be seen that the two storm

events of 1998 and 2012 are fairly similar in shape, except that the peak in the 2012 graph is

31% smaller than the 1998 graph (in reality the 1998 graph would be much larger due to the

under reporting problem). The cutout at the top right corner of Figure 10 expands the start of

the storm event and shows that the storm event of 2012 (2011 graph removed for clarity)

had a much more explosive impact. An analysis was carried out on the first 34 faults

reported on both of the storm events, this figure was selected as it is approximately where

each of the graphs start to rise in parallel. The storm event of 1998 reached 34 faults at

19:09 which was 14 hours 7 minutes after the initial fault, the storm event of 2012 reached

34 faults at 07:35 which was 2 hours 5 minutes after the initial fault was recorded.

Figure 10 – HV Fault Count

Figure 11(A) shows the amounts of customers off supply during the storm events. Again the

graph illustrates the number of customers effected at the start of the storm is greater at the

start of the 2012 event and rises more sharply than the storm event of 1998. Figure 11 (B)

illustrates the main timings of the two storm events. The 1998 storm reached its peak

number of customers (87,868) in 19 hours 35 minutes (2) and had a clear up duration of 7

days 18 hours and 37 minutes (4), this is compared to the 2012 event, which reached its

peak number of customers (72,082) in 4 hours 51 minutes (1) and its clear up duration

lasted 4 days 16 hours and 59 minutes (3).


Project No. 82880

15

Figure 11 – (A) Customer Interruptions Count (B) Storm Times

The HV network during the storm event of 2012 had the following characteristics compared to the storm event of 1998:

The fault (Damage) count was 31% lower than 1998.

The main damage happened 86% quicker than the storm of 1998.

The maximum number of customers off supply at any one point was 18% lower than 1998 (customer figures reported in the Boxing Day storm review report would suggest that this value is 42% lower than 1998).

The clear up duration was 39% quicker than 1998.

4.1.1 Network Failures

HV network faults are categorised as ‘Conductor’, ‘Insulators’, ‘Jumpers’, ‘Poles’, and

‘Others’. Figure 12 illustrates the fault distributions from each storm event, and shows a

proportional increase in Insulator faults.

Figure 12 – Fault distributions

Figure 13 shows that for the 2012 storm event compared to the storm event of 1998 there

was a 48% reduction in Conductor faults; 14% increase in Insulator faults; 53% reduction

in Jumper faults and a 43% reduction in Pole faults.

In the 1998 storm event 16% of the HV damage faults were attributed to trees and 84% were

classed as windborne damage. The panel of inquiry set up after the Boxing Day storm

reported [1] that the primary cause of 85% of all LV and HV faults was due to trees,

therefore the statistic of 85% tree damage will be used for 1998 in this document.

The reporting process and data handling during storm events has been improved since the

1998 storm. The 2012 data has 35%2 of the faults had no secondary cause compared to

77% of the 1998 data; therefore the reporting of the 2012 fault data should have a higher

degree of accuracy.

2 The data set was incomplete, as the secondary cause was not available for ‘Central & Fife’ and

‘Dumfries’ regions at the time of writing.


Project No. 82880

16

During the 2012 storm event 61% of the HV damage faults were attributed to trees and

39% to other factors, mainly windborne debris.

Figure 13 – Fault Categories

4.2 LV Network Performance

The LV network suffered from 6,051 LV faults during the 1998 storm event whilst in 2012 it

suffered from 518. This is a reduction in LV faults of 92%.

Figure 14 – LV Faults

After the 1998 storm event the panel of inquiry found that there had been a lack of focus on

the LV network, and that it appeared to be less well maintained than other networks. The

poles were also generally considered to be in a worse state than other voltage networks plus

coupled with the difficulty of obtaining permission to clear vegetation due to householders

wanting to shield the view of the overhead line, made the LV more prone to storm damage.


Project No. 82880

17

4.2.1 Network Failures

Figure 15 shows that ENA TS 43-30 [12] (LV open wire) suffered the highest proportion of

the 2012 storm event’s LV faults with The population of SPENs Scottish LV network is

approximately 95% ENA TS 43-30 and 5% ENA TS 43-12 [13] (ABC).

The total proportion of ABC faults is 3% (2% ABC Main, 1% ABC Service), which would

indicate that ABC has an increased resilience during storm damage over open wire

construction. The major component in the LV fault figures is conductor damage (Figure 12),

therefore as the population of ABC increases within SPEN’s Scottish LV network the LV fault

levels should further reduce during storm events.

Figure 15 – LV faults per construction

4.2.2 LGC Programme

Currently approximately 95% of Scottish Power’s LV network is constructed to ENA TS 43-

30 [12] construction specification. This specification entitled ‘Low voltage overhead lines on

wood poles’ uses wood pole supports and bare wire conductors attached to reel or screw-in

insulators. This specification is prone to clashing of the conductors due to their relatively

small spacing of 300mm. The main alternative to the 43-30 construction is ENA TS 43-12

[13] which uses Aerial Bundled Conductors, which due to their insulated nature do not have

the problems associated with clashing.

SPEN have now introduced a large LV modernisation programme (LGC Programme) and

plans to intervene on 1,000 km of LV every 5 years. This programme initially looks to

replace ENA TS 43-30 construction with ENA TS 43-12 or sections or underground cable.

Due to the cost of undergrounding overhead lines, the main component in the LGC

programme will be upgrading LV OHLs to ENA TS 43-12.


Project No. 82880

18

4.3 ABC

SPEN’s long term modernization strategy for LV overhead lines encompasses a prioritized

investment programme of Aerial Bundled Conductor (ABC) and selective undergrounding

[14]. The strategy states that no more bare wire LV OHLs will be added to SPEN’s LV

network.

Figure 16 – LV Constructions

Figure 16 illustrates the difference between the conductors design and ultimate working

tensions; all three of the ABC conductors show a very large extra capacity over its design

tension. This can sometimes be detrimental to an OHL as poles become the weakest link.

ABC has such tension capabilities that it has been witnessed were stays have pulled through

their concrete blocks and the pole snapped before the conductor failed. This can be

remedied by introducing weak parts in the suspension or tension fittings.

During HV refurbishments SPEN’s Manweb area has carried out re-conductoring of

the first span from OHL substations plus renewing the first pole if required. This was

carried out to reduce the need for a further HV shutdown during the refurbishment of

the LV network; however during storm conditions there has been a reduction in the

amount of HV faults caused by the failure of the first LV span.


Project No. 82880

19

4.4 Ofgem Metrics

In the DPCR5 price review, Ofgem specifies a number of outputs from each DNO that are

designed to define and measure a company’s performance. These are based on the

systems and information that is available in each company; however the areas of customer

interruptions (CI) and customers minutes lost (CML) will be reported as common metrics

across the industry.

The definition of CMLs and CIs is as per Ofgems definition [15]:

CML – The duration of interruptions to supply per year – average customer minutes lost per

customer per year, where an interruption of supply to customer(s) lasts for three minutes or

longer, calculated as:

The sum of the customer minutes lost for all restoration stages for all incidents The total number of customers

CI – The number of customers whose supplies have been interrupted per 100 customers per

year over all incidents, where an interruption of supply lasts for three minutes or longer,

excluding re-interruptions to the supply of customers previously interrupted during the same

incident. It is calculated as:

The sum of the number of customers interrupted for all incidents ∗ 100

The total number of customers

Table 4 – CML and CI results for the 03/01/2012 storm event

Table 5 displays the comparison of the 2012 storm events CI and CML results identified in

Table 4 when compared to 1998’s in Table 1 shows:

Table 5 – 1998/2012 CI and CML comparison table

A 64% reduction in the overall CMLs.

A 48% reduction in the overall CIs.


Voltage West

Scotland

East

Scotland

EN North West

Scotland

East

Scotland

EN North

EHV 2.30 0.02 2.32 0.52 0.10 0.63

HV 13.98 15.98 29.97 2.28 2.80 5.08

LV 1.09 0.91 1.99 0.10 0.09 0.19

Total 17.37 16.91 34.28 2.91 2.99 5.90


Voltage 1998 2012 Reduction 1998 2012 Reduction

EHV 8.71 2.32 73% 1.92 0.63 67%

HV 82.04 29.97 63% 9.15 5.08 44%

LV 3.36 1.99 41% 0.29 0.19 34%

Total 94.11 34.28 64% 11.36 5.90 48%


Project No. 82880

20

5 Improvement Policies

5.1 OHL Policies

OHL-01-008 [3] was introduced in 2002 in response to the storm event of 1998. It was

introduced to improve network resilience by addressing the inadequacy of lightweight

constructions such as BS1320. This was particularly pertinent to main lines that were built

during the electrification of rural areas in the 1950’s and 60’s.

OHL-03-099 [16] lays out Scottish Power’s specifications for all new or re-build 11kV and

33kV OHL circuits. All of the specifications are of unearthed construction with AAAC

Conductors, supported on single and “H” type wood pole supports. OHL-03-099 has nine

construction specifications (Table 6), each of which are categorised as severe and non-

severe. OHL-01-008 specifies the heavier construction standard of L27, for HV circuit main

lines being constructed or re-built in the defined Severe Weather area. It also states that a

HV spur less than five spans in length can be constructed to the L27 specification if

connected solidly to the main line.

The heavier L27 specification is designed to be resilient in severe weather area conditions;

however there is anecdotal evidence that L10 lines withstood the 2012 storm event within

the defined severe weather areas.

Table 6 – OHL construction designs (OHL-03-099) Construction Conductor Voltage Conductor Size Weather Area

All Aluminium Alloy Conductors (AAAC)

L10 Hazel 11kV 50mm AAAC Non-Severe

L27 Oak 11kV 100mm AAAC Severe

L27 Ash 33kV 150mm AAAC Non-Severe

L33 Poplar 33kV 200mm AAAC Non-Severe

L33 Poplar 33kV 200mm AAAC Severe

L35 Upas 33kV 300mm AAAC Non-Severe

Optical Phase Conductors (OPPC)

L34 Poplar equivalent conductor 33kV 200mm OPPC Non-Severe

L34 Poplar equivalent conductor 33kV 200mm OPPC Severe

L36 Upas equivalent conductor 33kV 300mm OPPC Severe

Each of the OHL designs has a basic recommended span, which when plotted against the

conductors maximum span and clashing factor as per ENA TS 43-40 can illustrate how likely

the particular OHL design’s conductors are likely to clash.


Project No. 82880

21

Figure 17 – OHL: Basic – Maximum Spans

Figure 17 shows the difference between an OHL’s basic recommended span and its

maximum span, which will avoid clashing dependent on the designs conductor spacing

(crossarm size). The maximum span was calculated for an ENA TS 43-40 2C weather area,

which is defined as the OHL would be subjected to a wind pressure of 380N/m2 and a

diametric ice acretion of 30mm. It shows that BS1320 and ENA TS 43-10 constructed lines

would be prone to clashing as their basic recommended spans are greater than the 43-40

calculated maximum span. Prior to 2007, SPEN constructed its L10 lines to the Severe

weather appendix of ENA TS 43-10. In 2007 L10 was re-written to an increased

specification. Figure 17 also displays the 43-40 clashing factor, which for display purposes

has been factored by 100.

Figure 18 – OHL Conductors

Lines designed in accordance with ENA TS 43-40 are more likely to withstand the likely wind

and ice loadings as identified within the relevant UK weather zone maps. Figure 18 shows

the conductor failure curves against the wind and ice loadings for span lengths equal to the

basic/recommeded and maximum span lengths. These findings have been replicated from

the STP project S0283 [17]. The conductor failure curves display the maximum conductor

load that can theoretically be applied to the conductor. It illustrates the load at which the

Maximum Spans Basic/Recommended Spans


Project No. 82880

22

conductor would fail either by reaching its elastic limit or by exceeding the conductors actual

ultimate tensile strength (UTS). The graphs have the 43-40 weather areas overlaid in red,

for extra clarity 380N/m2 is equal to 55.7mph, 570N/m2 is equal to 68mph and 760N/m2 is

equivalent to 78.9mph. Upas is only displayed in the maximum span length graph as the

L35 specification recommended span length of 80m greatly out performs the programme

(used for calculations) limit of 180mm diametric ice acretion.

Figure 18 identifies that Hazel conductor ‘maximum spans’ can withstand 1E, 2E, 3D,

4C and 5B ENA TS 43-40 weather co-ordinates. SPEN rate their L10 specification

(Hazel conductor) to a maximum of 2D & 3C weather co-ordinates, and anecdotal

evidence suggests that it performed very well during the 2012 storm event.

During the storm event of 2012, 4 permanent faults occurred on main lines that had

been upgraded by the improvement policies and one on an OHL that had been

refurbished. 713 km of HV OHLs have been upgraded so far out of an asset

population of 14,048km, which equates to 5% of the HV OHL population. The fault rate

of upgraded mainlines is 1%, this illustrates an overall improvement in HV mainline

resiliance due to the proportion of faults being significantly lower than the proportion

of population.

There was no permanent faults on any refurbished Spur lines during the 2012 storm

event.

5.2 Vegetation Management

The first improvement programme implemented post 1998 was ‘Rural Care’, which saw a

change from vegetation clearance on a cyclic basis to one that concentrated on tree

clearance for vegetation below or within falling distance of HV overhead lines. This policy

was applied to OHL construction and refurbishment projects plus circuits identified as poorly

performing. This policy attempted to clear all trees identified within the proximity zone 1

defined within ENA ER G55 [18], in reality this policy was not liked by landowners and the

general public and consents were hard to get approved especially within TPO (Tree

Preservation Order) areas.

The introduction of OHL-01-008 in 2002 introduced Rural Care 2, which was a more

measured approach to vegetation management by considering vegetation control within high

and low risk zones. Zone 1 maintains a 2.0 metre clearance zone around HV lines and Zone

2 requires high risk trees to be cleared within falling distance of HV lines. Arborists give

guidance for the condition of certain tree species within Zone 2. Rural Care 2 is applied to

all overhead line rebuild and refurbishment projects.

In 2006 the Energy Networks Association (ENA) issued ENA ETR 132 [2] which saw a move

towards “a risk based methodology that provides guidance on how to proactively improve the

overall performance of a Network by improving the Resilience of the Network to Vegetation

related faults that can occur under Abnormal Weather conditions”. ETR 132 recognises the

fact that policies such as rural care are unacceptable, on an environmental basis; however a

risk will remain under abnormal storm conditions that healthy trees might still fall causing

Network faults. There is anecdotal evidence that during the storm event of 2012 trees that

would have been regarded as healthy had in fact fallen causing a permanent fault.


Project No. 82880

23

Figure 19 illustrates the monthly rainfall figures for 2009 to 2011 plus the mean average from

2001 to 2010. The 2011 graph shows above average rainfall figures for most of the year,

the December figures show a massive increase above average. It would be safe to assume

that the significant increase in rainfall had contributed to some tree foundation instability,

contributing to some of the healthy tree failures.

Figure 19 – Monthly Rainfall

Climate change is predicted to bring about climatic changes that will produce more extreme

storms, which will be classed as exceptional events or conditions. A Met Office report [19]

predicts that the UK will experience increasingly adverse climatic conditions over the coming

decades. DNOs will therefore require increasing levels of investment due to the increased

difficulty in maintaining Network Resilience, including Vegetation Management.

The World Meteorological Organization (WMO) defined "Abnormal Weather" as a climatic

phenomenon that occurs only every 25 years or more. Since 1998 SPEN’s Scottish network

has been subjected to 10 storm events with maximum wind gusts in excess of 70mph plus

severe snow storms in early 2001 and March 2010. In date order the severe weather events

were:

“Snowblitz” – February 2001 (Snow storm);

“Cyclone Noah” – 1 January 2002;

“Cyclone Erwin” – 8 January 2005;

“Cyclone Gero” – 11 January 2005;

“Cyclone Franz” – 10 January 2007;

“Hungry Snout” – March 2010 (Snow Storm);

“Hurricane Katia” – 11 September 2011;

“Cyclone Xaver” – 25 November 2011;

“Cyclone Friedhelm” – 8 December 2011;

“Cyclone Hergen” – 11 December 2011;

“Cyclone Ulli” – 3 January 2012;

“Cyclone Andrea” – 4 January 2012.

It can be observed that six storm events occurred in the 2011/2012 winter period and four of

these were within one month at the end of 2011 and start of 2012. These would substantiate

the Met Office’s prediction of increased adverse climactic conditions.


Project No. 82880

24

SPEN originally operated a five year cyclical vegetation clearance programme; this has now

been reduced to a three year programme. The vegetation programme has been reduced to

a three yearly cycle as SPEN does not believe that clearances as specified in ENA TS 43-8

[8] can be achieved with a cycle greater than 3 years. This is largely due to the increased

growth rate of vegetation and the restricted cuts enforce by landowners.

During re-build projects it has sometimes been difficult obtaining compliance with

ETR 132 due to small areas of excessive vegetation or reluctant landowners. As ETR

132 compliance has to be reported on the whole line this could become a factor in not

refurbishing an OHL.

During the 2012 storm no ETR 132 trees fell causing any permanent faults.

5.3 Severe Weather Map

Scottish Power with assistance from the Met Office developed a severe weather map for the

Scottish network (Figure 20). The severe weather map illustrates two distinct weather areas:

Severe (Blue) – with mean hourly wind speeds up to 70mph, gusts up to 105mph.

Normal – with mean hourly wind speeds up to 50mph, gusts up to 75mph.

Figure 20 – Severe Weather Map

Figure 21 illustrates the UK weather zone maps from ENA TS 43-40, which were created to

define diametric ice and mean wind pressure co-ordinates for sites between 0m and 500m.

The maps from left to right are 0 to 100m; 100m to 200m; 200m to 300m; 300m to 400m;

400m to 500m.

Figure 21 – 43-40 Weather Zone Maps

STRANRAER

NEWTON STEWART

NEW GALLOWAY

DALBEATTIE

DUMFRIE

S

LOCKERBIE

LANGHOL

M

MOFFAT

HAWICK

KELSO GALASHIEL

S

PEEBLES

LANARK

HADDINGTON

DUNBAR

LOANHEAD PENICUIK

DUNFERMLINE

CUMBERNAU

LD

KILSYTH

STIRLING

KILMARNOCK

STEWARTON

SANQUHAR

EAST KILBRIDE

CUMNOCK MAYBOL

E


Project No. 82880

25

The comparison between the ENA TS 43-40 weather zone maps and SPEN’s severe

weather map illustrates that SPEN’s severe weather map would benefit from some

refinement.

Figure 22 illustrates the storm event’s HV fault distribution for each region divided in to their

respective weather areas. 79% of the faults occurred within non-severe weather areas and

19% occurred within severe weather areas. The distribution of faults that occurred within

severe weather areas were:

Borders – 46%

Dumfries – 12%

Lanarkshire – 42%

Figure 22 – Weather zone faults

97% of the HV faults in the Borders region occurred within a severe weather area; 59% of

Lanarkshire region’s HV faults occurred within a severe weather area and the Dumfries

region had 15% of their HV faults occur within a severe weather area.

The severe weather map (Figure 20) identifies that:

Ayrshire & Clyde South has a small severe weather area (Approximately the bottom

half of the small blue area above Stewarton (Figure 20)). All of Ayrshire & Clyde

South region’s 77 faults occurred within the non-severe weather areas. Although the

severe weather area is quite small it would have been expected, given the large

number of faults for this region, that some damage would have occurred within it.


Project No. 82880

26

The Borders region is predominantly a severe weather area, which is reflected in the

fact that 97% of the faults were classified as occurring within a severe weather area.

The Central & Fife region has a severe weather area (above Kilsyth (Figure 20)),

which accounts for approximately 20% of the region. This region suffered the

largest amount of HV faults; however none had occurred within the severe weather

area. It would have been expected to see damage within the severe weather area

given its size and the number of faults report in this region.

Approximately 66% of the Dumfries region is a non-severe weather area, and 85% of

all of the faults that occurred in this region were classified as occurring within the

non-severe area.

The Edinburgh & Lothians region also has approximately 66% of its area defined as

non-severe; however all of its faults were categorised as occurring in this area.

The Glasgow & Clyde North region has a small severe weather area (Approximately

the top half of the small blue area above Stewarton (Figure 20)). Given the small

severe weather area and the relatively small amount of HV faults that occurred in

this region, no conclusion can be drawn from the fact that all of the faults occurred

within the non-severe area.

The majority of the Lanarkshire region is classified as a severe weather area and

59% of the HV faults were reported as occurring within this area.

As Figure 22 illustrates the Ayrshire & Clyde South, Central & Fife and Edinburgh & Lothians

regions’ HV faults were all classified as occurring within non-severe weather areas; however

Figure 20 identifies severe weather areas within these regions.

A number of reasons could explain the lack of severe weather area HV faults within the

regions identified above:

The storm passed below the severe weather areas. The storm did pass through all

of these areas, as there is anecdotal evidence that the storm passed through the

Loch Lomond area, which is north of all of these regions. There is also evidence

that SSE suffered OHL faults across the central belt of Scotland.

The OHLs within the severe weather areas could have received a higher percentage

of improvement than other regions, or the population of OHLs in these regions are

lower than other regions.

The data regarding the weather regions could have been incorrect.

An investigation into why the Ayrshire & Clyde South, Central & Fife and Edinburgh &

Lothians region’s severe weather areas performed so well compared to other areas

could identify that the severe weather map requires some refinement or some other

reason that could benefit the other regions.


Project No. 82880

27

6 OHL Protection Policy

Remote areas and areas of low population density are usually served by OHLs; restoration

of supplies to these areas can sometimes be delayed by difficult terrain and longer distances

between DNO depots and customers. Also during storm events these areas will be given a

lower priority over higher population density circuits. 50% of faults on overhead lines are

transient faults, which can have many causes, including debris being blown onto the lines,

birds or lightning. The effect of a transient fault on a standard OHL would trip the protection;

the fault would then disappear (usually by being blown away by the power of the initial fault).

The OHL would therefore be in a serviceable condition and off supply; SPEN introduced the

overhead line protection policy (OHPP) in 1993/94 to overcome these issues.

The OHPP uses microprocessor-controlled OHL switchgear that disconnects faulty

equipment as quickly as possible. After a pre-set time the switchgear will automatically

attempt to restore supplies, if the fault was transient then supplies would be restored. The

customers connected to the OHL during transient faults would only experience a short

disconnection compared to a lengthy one.

Originally due to their design automatic switchgear would have a number of operations

before they would ‘lock out’, resulting in a loss of supply. Although this would result in a

lengthy loss of supply it would alert SPEN to the fact that there is a potential problem with

the affected circuit. Since the inception of the OHPP automatic switchgear can now carry

out many thousands of operations without maintenance; therefore they can be set to

unlimited operations. SPEN’s automatic switchgear has a reclose time of 10 seconds and

reclaim time of 10 seconds. If a fault is detected within the reclaim time the switchgear will

proceed through its 3 shot cycle every time a fault is detected before the reclaim time; after 3

cycles if lock out indicating a permanent fault. If the circuit stays on supply for longer than

the reclaim time the switchgear will reset back to zero shots.

Spur lines under OHPP are protected by pole-mounted automatic sectionalisers. These

discriminate between a transient and persistent fault by counting the passage of fault current

during the auto-reclose sequence of the controlling switchgear. Sectionalisers operate

during the dead time of the auto-reclose sequence after a pre-determined number of

passages of fault current. They are only fitted to circuits protected by a multi-shot auto-

recloser with minimum number of trips to lock-out being one more than the sectionaliser

count.

Many overhead line faults are transient in nature due, for example, to wind activity or

lightning. The application of the above overhead line protection policy, introduced in

1993/94 improves system performance under transient fault conditions and limits the number

of customers disconnected under persistent fault conditions [14].

There is some anecdotal evidence to suggest that some OHPP equipment did not perform in

the way that it was designed during the recent storm event. These were mainly thought to

be primarily sectionaliser links and failures of GVR batteries. Sectionaliser links are

periodically tested and during these tests earlier versions with known problems are routinely

replaced.


Project No. 82880

28

The GVR batteries have a 10 year life expectancy and are checked at 5 years, some failures

are detected at this time. It is thought however that approximately the same percentage of

failures were experienced during the storm event as would be expected during a standard

inspection.

There is also anecdotal evidence that some customers have suffered from many short

interruptions (NaFIRS description [5]), and SPEN are only aware of the problem once

a complaint is made. Other UK DNOs have applied the automatic switchgear at

primary substations, this has the advantage of SCADA communications but the

downside is, more customers being affected and counted on each short interruption.

As rural pole mounted automatic switchgear does not usually have any SCADA

communications, devices can be employed in customer’s premises, streetlights etc. that can

communicate with the control office and register that they have experienced a loss of supply.

This would alert SPEN to persistent transient faults before customers start to complain.


Project No. 82880

29

7 Poles

7.1 Installation

The two main excavation methods for pole installations are with an excavator or auger. All

of Scottish Power’s HV OHL constructions employ pole baulks on all of their poles; therefore

no HV poles can be installed with the auger method. Some other UK DNOs allow HV poles

to be installed without pole baulks as long as the poles are installed 0.5m deeper than

normal.

As LV poles do not have any pole baulks installed, LV poles can be installed with the

excavator or pole auger methods. Installing poles with the auger method minimises the

amount of ground that needs excavating, however augered holes are usually so tight to the

pole that no reinstating can be carried out (Figure 23A) or so large that too much

reinstatement is required (Figure 23B). In other words no reinstatement material can be

packed down the sides of a tightly augered hole and too much ground is removed on large

augered holes. Both of these scenarios result in leaning poles.

Poles installed with an excavator are usually easier to obtain a tight hole as it can be

reinstated through the corners (Figure 24). Both of these pole installation methods are valid,

however the quality of the pole’s foundation is directly attributable to the quality of the

installation. Therefore auditing of pole installations or holding contractors to account for

leaning poles for at least 2 years will increase the quality of pole installations.

There is no evidence of pole failures due to poor foundations during the storm event

of 2012; however these failures can go undetected, as leaning poles do not always

threaten the security of the supply.

Figure 23 – Augered pole holes

(A) (B)

Figure 24 – Excavator pole hole


Project No. 82880

30

7.2 Condition

SPEN currently carry out AIS inspections every 6 years pole poles categorised as normal

risk, high risk poles are inspected on a more regular 3 year basis. Figure 25 illustrates the

timeline of inspections for OHL circuits. It shows that inspections are carried out for the

vegetation management programme every 3 years, OHL inspections are carried out every 6

years (3 years for high risk poles). The OHLs are refurbished every 12 years (8% of

population), the interim OHL inspection between refurbishments are to catch any defects

between refurbishments. No decayed poles will be expected during these inspections as

inspectors are asked to assess the poles on Refurbishment inspections for a minimum of 12

years life. OHLs will be evaluated at the 12 year period for refurbishment, 1% of the circuits

would be selected for re-building.

Figure 25 – Inspection Timeline

Inspectors carrying out OHL inspections evaluate a poles condition through visual inspection

and a hammer test (Figure 26). The pole’s condition is then reported in the AIS system as

OK or defective. If SPEN encounter a larger than usual amount of defective poles, a further

evaluation is carried out by an outside contractor to ascertain the levels of decay using more

advanced pole testing methods.

When poles are evaluated with such a basic test as a hammer test poles that are in really

good, normal or really poor condition are easily identified. Poles that the inspector is not

sure about, end up being classed as defective just to be on the safe side. If the AIS pole

evaluation was categorised into 4 levels such as ‘New’, ‘OK Normal Wear’, ‘Poor’ and ‘Bad

Work Required’, the poles that are not easily determined with a hammer test would fall into

the ‘Poor’ category and could be tested further. Some UK DNOs ask their inspection

contractor to carry out the further test on site with equipment that can easily be carried to the

pole such as a ‘Matson Corer’ or ‘Purl Tester’.


Project No. 82880

31

Figure 26 – Hammer Test

Whilst pole testing using the ‘Hammer Test’ is very subjective using a condition rating should

improve the quality of pole testing. It can also increase the ability to audit an inspectors work

as level 3 poles (‘Poor’) should never be reported as any other level, just as level 4 poles

(‘Bad, Work Required’) should only ever be reported as level 4.

One common cause of missed diagnosis of a pole is that the inspector does not carry out a

full hammer test. The test should be carried out from the ground level right up to full height

with extend arm. One UK DNO experienced a pole failure after a pole had been assessed

for pole rot and categorised as ‘OK,Normal Wear’, this was later found to be because the

inspector was not bothering to bend down and hit the pole at the ground line. Simple

auditing carried out after an inspection can detect if vegetation has been cleared and or

whether hammer marks extend down to the ground line.

There is no evidence that any poles failed solely through pole decay during the 2012

storm event. One pole was described as having pole decay; however it is not explicit

that this was the primary cause of the failure.


Project No. 82880

32

8 Targeted Solutions

Predominantly SPEN’s HV network is constructed of bare wire horizontal construction

standards with a small amount of old historic specifications such as Genu 3. SPEN also has

a small amount of covered conductor (approx. 6km), which accounts for 0.04% of its HV

network. The use of a targeted solution could improve the performance of some OHL

circuits that currently are experiencing difficulties due to harsh or difficult

environments that are not conducive to bare wire construction. For example OHLs

that have a poorly performing section due to bird strikes, could have this section re-

placed with a more visible and possibly insulated conductor. The performance of the

network during the 3rd January 2012 storm event has shown that wholesale

employment of alternative OHL constructions is not required, however the recognition

and approval of alternative construction/s would allow SPEN more flexibility when

encountering problem areas. It is recognised that having small quantities of

alternative OHL constructions can present problems during maintenance and fault

work, due to the workforce having very little experience of the constructions. SPEN

currently has small quantities of BLX and Ericsson OHL construction, which was

constructed by a previous OHL contractor. This means that SPEN and its current

OHL contractor has very little experience these constructions; and therefore it would

be prudent to train some of SPEN’s workforce in these constructions. There is also a

need for some material and equipment to be purchased for maintenance and fault

work to be carried out.

The targeted solution could also allow OHLs to be built with reduced tree cutting, where ETR

132 tree cutting has been restricted or refused due to landowner issues.

Detailed below are a sample of constructions that could be targeted at problem areas, the

use of targeted solutions would give SPEN a greater ability to match the OHL construction to

its environment.

8.1 BLX

BLX is taken from the Norwegian 'Belagt Linesystem', or overhead line system, insulated

with XLPE. The line consists of a stranded, alloyed aluminium conductor with cross-section

from 50 mm2 to 150 mm2, covered with a cross-linked and UV resistant polymer. The

conductor is grease filled during the manufacturing process to make the covered cable

watertight and corrosion proof.

One of the main benefits of BLX is that unlike conventional bare wire OHLs the phases can

be suspended very close together (Figure 27). The smaller crossarm size allows for a much

reduced wayleave than for a standard OHL, plus it would apply much less torque to the pole

top if it or the conductors were hit by a fallen tree or windblown object due to its smaller size.

In an early pilot study in Norway, trees were deliberately felled at their roots in order to fall

onto the BLX line. The line remained live throughout the study and regular inspections were

carried out to investigate any damage. After a year the study revealed no signs of

mechanical wear on the covering of the line.


Project No. 82880

33

Experience with BLX over lakes and rivers in landing/ascent areas for swans and other

water birds have also been very positive. Seen against the sky, the line is both more visible

and compact than bare line and the birds in flight seem better able to avoid it.

The working practices associated with building conventional bare line can easily be applied

to BLX installations. The covered conductors can be drawn through rollers and guided so

that the covering is not damaged.

There is anecdotal evidence that during fault conditions BLX has been reconductored with

bare conductor due to insufficient stocks and knowledge of the construction. The use of

bare conductor on BLX crossarms subsequently resulted in clashing problems.

Figure 27 - BLX

8.2 Ericsson

Ericsson cable is a multi-purpose cable that can be installed in different environments and

conditions from being installed as an underground cable, submersed in water and

suspended as an OHL. The cable has to be robust enough for laying underground; have

adequate density that it will sink for laying in water; possess self-damping properties and be

able to withstand the sometimes extreme conditions a self-supporting aerial cable can be

exposed to from ice and wind loadings.

The internal design of the cable affects the way vibrations and galloping are damped by the

cable. At EA Technology’s test field on the Shetlands the Ericsson cables were installed on

spans up to 90 meters for a period of 18 months [20]. The cables were monitored by load

cells and video cameras and performed well under weather conditions with wind speeds (10

minutes average) of up to 82 knots (94 mph). Ericsson cables were further tested at EA

Technologies Deadwater Fell test site [21].

If a falling tree lands on an Ericsson OHL, there is a risk that the cable could fail if the

tree falls close to a terminal or section pole. If it however fell in the middle of the OHL

then the cable will stay intact by slipping through the suspension clamps; there could

also be some elongation of the cable. After removal of the tree the line would require re-

regulating and any elongation removed through the nearest termination.


Project No. 82880

34

Figure 28 – Ericsson Cable (Mull of Galloway Lighthouse)

8.3 Spacer Cable Systems

Spacer cable systems consist of a tensioned messenger wire under which is supported,

three covered conductors (CC) in a polymeric support (spacer) cradle. The CC phase

conductors are configured in a close triangular formation and can be double or triple

sheathed. The Spacer cable system has high mechanical strength to weather severe

storms, and good electrical strength that prevents phase to phase and phase to earth faults

plus animal and tree contact [22].

The messenger wire provides a high strength physical shield against falling branches

and trees. Its compact nature and its ability to withstand temporary tree contact makes it

a good candidate for areas where cost or refused permission means only a reduced tree

clearance can be obtained.


Project No. 82880

35

9 OHL Maintenance

SPEN currently refurbishes 8% of its OHL circuits and re-builds 1% on an annual basis. The average

cost of re-building a km of OHL is 10 times, the cost of refurbishment.

Circuits are identified for refurbishment or re-building by a formula that weights ‘decayed/damaged

poles’, ‘conductor faults’, ‘stay faults’, ‘insulator faults’, ‘customer numbers’ and ‘fault rate’ to rate the

circuits in order of priority. The top 8% of circuits would be selected and 7% selected for

refurbishment and 1% would be selected for the re-build list.

SPEN has an obligation to deliver the refurbishment and rebuild quantities as part of a DPCR5 output,

circuits selected for refurbishment that are subsequently found to cost vastly more than the

refurbishment budget are placed on the re-build list. Once a circuit reaches the top of the re-build list

it would take a minimum of 2 years to progress it from start to actually re-building the OHL largely due

to landowner issues.

As part of this network review process a 33kV pole was identified as having failed in one of the recent

storms on a line that was thought to have been refurbished. It was decided that this would be

investigated to see why it had failed. During the investigation it became clear that the circuit was due

to be refurbished, however due to cost of replacing a large number of rotten poles on the circuit, it was

placed on the rebuild list. This circuit had been prioritised and it is due to be re-built in 2012/13.

Currently the choice is to refurbish or re-build a circuit, however there is no category for

circuits that require urgent attention but would cost more than the refurbishment budget per

km, except to be classed as rebuild. An interim choice would allow OHLs to have a degree of

re-build to the older specification of the original line i.e. original span lengths.

9.1 Prioritisation

9.1.1 HV

Figure 29 illustrates the amount of faults per population of connected customers, it shows that of the

20 faults with more than 2000 connected customers 17 (85%) are HV main Line, 2 (10%) are EHV

and 1 (5%) is a HV spur line. Whilst the spur line only represents a small percentage of the faults with

more than 2,000 connected customers it would be prioritised within the storm clear up work due to its

number of customers. Therefore spurs with large numbers of connected customers should be

considered for inclusion within the improvement programmes and be controlled by its own

auto-recloser.

Figure 29 – Faults per population


Project No. 82880

36

9.1.2 LV

Figure 30 shows that the majority of all faults from the 2012 storm event occurred on the LV

network. Figure 29 illustrates that 1 (4%) of the greater than 500 customer connected faults

was an LV fault, however this is thought to be a miss-reported piece of data, as it is doubtful

that SPEN has an LV OHL with more than 500 connected customers, as SPEN restrict their

LV OHLs to 200amps per phase.

11 LV OHLs with permanent faults had between 50 and 300 connected customers these

circuits would have been quite highly prioritised during the fault repair stage of the storm

event.

Figure 30 – Fault distribution

The majority of the LV faults are, as would be expected in the less than 20 connected

customer’s category. As the customer interruption count curve shows in Figure 31,

the tail end reduces a lot slower than the start of the storm event; this is due to the

amount of less than 20 connected customer faults. One metric that is important to a

DNOs customers after a storm event, is being off supply for longer than 48 hours.

During the storm event of 2012 4,792 customers were off supply for more than 48

hours, this is a reduction of 66% over the 1998 storm.

Figure 31 – Customer Count


Project No. 82880

37

10 Conclusions

The storm event of the 3rd January 2012 was comparable to the storm event of the

26th December 1998.

The storm event of 2012 had an 87% reduction in total faults compared to the 1998

storm event.

The storm event of 2012 reached its peak number of customers interruptions 31%

quicker than 1998.

The maximum number of customers off supply at any one point was 18% lower in

2012 than 1998.

The 2012 storm clear up duration was 39% quicker than 1998

From a diminishing number of faults from 1998 to 2012, ‘Insulator’ faults rose by

number and therefore proportion.

The major cause of damage was attributed to trees at 61% in 2012 compared to 85%

in 1998.

The CML result for the storm event of 2012 was 64% lower than 1998

The CI result for the storm event of 2012 was 48% lower than 1998.

The resilience of the HV and LV networks during the 2012 storm event has

dramatically improved compared to the storm event of 1998; however the same level

of improvements has not been reflected in the CML and CI values for the LV

Network. As the LV CML reduction is similar to the LV CI reduction the problem

would indicate that more customers were effected per fault

The L10 OHL construction performed well under storm conditions.

According to the STP2 project S0283, Hazel conductor ‘maximum spans’ can

withstand 1E, 2E, 3D, 4C and 5B ENA TS 43-40 weather co-ordinates. SPENs L10

specification is rated by SPEN to a maximum of 2D & 3C, these conservative design

restrictions have been proved to work, due to the excellent performance of L10 OHLs

during the 2012 storm event.

Upgraded HV main lines suffered 1% of HV faults and upgraded spur lines suffered

no faults.

Some trees identified as healthy failed during the 2012 storm event.

Obtaining ETR 132 compliance is not always possible to attain for a complete OHL

circuit.


Project No. 82880

38

The severe weather map is not as detailed as the ENA TS 43-40 weather maps.

No HV faults occurred within the Ayrshire & Clyde South, Central & Fife and

Edinburgh & Lothians region’s severe weather areas.

Adding condition levels to the pole testing in AIS will improve the quality of inspection

data for the pole asset class.

All new or re-build LV OHLs should be constructed to OHL-03-101.

The renewing of all the first LV poles and spans to ENA TS 43-12 during HV

refurbishment / Re-build work increases the HV network resilience.

There is insufficient knowledge of alternative constructions such as BLX, Ericsson or

Spacer Cable systems.

During the 2012 storm, no ETR 132 trees fell, caused any permanent faults. The

increased use of ETR 132 should be considered before having small amounts of

alternative contructions on the network.

Some circuits are taken from the Refurbishment list and placed on the Re-build list

due to budgetary and OHL design restrictions.

An increase in LV OHL re-builds (LGC) should shorten the Storm clear up duration

by the number of small LV OHL faults.

Spurs with large numbers of connected customers should be considered for inclusion

within the improvement programmes.


Project No. 82880

39

11 Recommendations

The upgrading of OHLs and the carrying out of tree clearance to ENA TS 43-8 and ENA

ETR 132 has had an impact on the amount of faults over the duration of a comparable storm

event. The upgrading of OHLs and the frequency of tree clearance should be continued as

per SPEN’s current strategy.

An assessment should be carried out into whether the L10 construction could be utilised

within the current severe weather areas, thus allowing more OHL circuits to be rebuilt at a

lower cost; and allowing a more targeted approach to the use of L27. Further work should

also be undertaken to refine the severe weather map for SPEN’s Scottish network. The

STP2 project S2641 is undertaking work to develop new wind/ice load maps, which are

expected to be less onerous than those in ENA TS 43-40. The refinement of the severe

weather map taking into account the findings of S2641 could also more accurately target the

correct construction for the environment.

The S2541 project is a large project which may form a revision of the current UK wind and

ice loads. It will use details of major wind/ice incidents dating back to 1870, along with UK

Met Office and Deadwater Fell data to develop rime ice, wet snow and extreme wind

models.

Consideration should be given into an insulated conductor construction, as a targeted

solution for poorly performing circuits; or ones that are experiencing difficulties in obtaining

permissions for a re-build and/or ETR 132 compliance.

For circuits that will struggle to achieve ETR 132 compliance, consideration should be given

into declaring a percentage of the line as ETR 132 compliant; or dividing the circuit with ABS

switches either side of the anomoly and declaring either side as ETR 132 compliant and the

middle as non-compliant.

Pole condition should be added to the AIS inspections, as this will reduce the amount of

poles that need re-testing plus it will give a better understanding of the Health of the pole

asset class. As Ofgem move into the next price review (RIIO ED1) greater emphasis will be

placed on a risk-removed output measure for asset classes. DNOs will be required to take a

holistic risk-based approach to asset management, by reporting the risk of an asset class at

the beginning of the ED1 price review and, with investment, its risk will be at the end of ED1.

The important outcome at the end of the RIIO ED1 price review will be the Health and

subsequent risk of an asset class rather than how many kilometers of line were re-built. Risk

management processes like EA Technologies CBRM (Condition Based Risk Management)

process can evaluate the risk of an asset class through age, environment, probability of

failure and criticality, however it is much more accurate if condition data is also supplied.

Major investment is required in the re-building of ENA TS 43-30, it was identified in the

Boxing Day storm review report [1] as being in need of refurbishment after the storm event

of 1998. The investment will help reduce the amount of LV faults during storm events plus it

will drastically reduce the clear up duration and non-planned CMLs; and therefore the

amount of customers that experience supply interuptions in excess of 48 hours.


Project No. 82880

40

Although the re-building of OHLs with ENA ETR 132 tree cutting is the prefered method of

upgrading; the introduction of alternative OHL designs should be considered for areas of

restricted ETR 132 compliance or other environmental anomalies. Training will be required

for all of SPEN’s workforce that will come into contact with any alternative constructions, plus

stocks of materials and equipment should be made available for any maintenance.

An investigation into why the Ayrshire & Clyde South, Central & Fife and Edinburgh &

Lothians region’s severe weather areas performed so well compared to other areas. This

could identify that the severe weather map requires some refinement or some other reason

that could benefit the other regions.

During HV refurbishment consideration should be given to renewing the first LV poles from

OHL substations and replacing the first span with ABC conductor. This should reduce the

number of LV faults that escalate into HV faults, plus it will reduce the need for subsequent

HV shutdowns during LV refurbishment, which in turn will help reduce ‘Planned CIs’.

Spurs with large numbers of connected customers should be considered for inclusion within

the improvement programmes.

As HV ‘Insulator’ faults increased proportionally during the 2012 storm event it would be

prudent to investigate if SPEN’s network received more windborne damage or whether there

is an underlying problem with insulators. SPEN however are currently in the process of

issuing a new insulator tender, which will specify that all new tension insulators are

manufactured from silicone and, HV pin insulators will be manufactured from polyethylene.

Therefore there is very little value in such an investigation.


Project No. 82880

41

12 References

[1] J Menzies, (1999), ‘Report of power systems panel of inquiry into the impact of the

storms during 26th & 27th December 1998’, Scottish Power Energy Networks.

[2] ENA ETR 132,(2006), ‘Improving network performance under abnormal weather conditions by use of a risk based approach to vegetation management near electric overhead lines’, Energy Network Association.

[3] OHL-01-008, (2002),’ Construction Policy for Distribution Overhead Lines’, Scottish

Power Energy Networks.

[4] BS1320, (1946), ‘High-voltage overhead lines on wood poles for line voltages up to and including 11kV’, British Standards.

[5] ENA ER G43-3, (2002), ‘Instructions for Reporting to the National Fault Interruption Reporting Scheme’, Energy Network Association.

[6] ‘Electricity Safety, Quality and Continuity Regulations 2002’, UK Statutory

Instruments.

[7] ‘Electricity Safety, Quality and Continuity (Amendment) Regulations 2006’, UK

Statutory Instruments.

[8] ENA TS 43-8, (2004), ‘Overhead Line Clearances’, Energy Network Association.

[9] (2012), ‘Weather Review Jan 3 2012 ScotPower’, MeteoGroup.

[10] http://www.metoffice.gov.uk/climate/uk/interesting/2011_decwind/ accessed

17/02/2012

[11] I Wallace, (2007), ‘An Assessment of HV Overhead Asset Storm Resilience’, Kema

Limited.

[12] ENA TS 43-30 (), ‘Low voltage overhead lines on wood poles’ Energy Network Association.

[13] ENATS 43-12, (), ‘Insulated aerial bundled conductors erection requirements for LV

overhead distribution systems’, Energy Network Association.

[14] (2010), ‘Distribution long term development statement for the years 2010/11 to 2014/15’, Scottish Power Distribution.

[15] Electricity Distribution Price Control Review Policy Paper http://www.ofgem.gov.uk/

accessed 23/02/2012.

[16] OHL-03-099, (2007), ‘Specification for 11kV and 33kV Overhead Lines of Unearthed

Construction on Single and “H” Type Wood Poles’, Scottish Power Energy Networks.

[17] Wareing J B & Horsman S, (2003), ‘The performance and suitability of UK OHL design for snow and ice conditions in light of recent storms’, EA Technology Report

5574, S0283.

http://www.metoffice.gov.uk/climate/uk/interesting/2011_decwind/

http://www.ofgem.gov.uk/


Project No. 82880

42

[18] ENA ER G55, (2008), Safe tree working in proximity to overhead electric lines,

Energy Network Association. [19] (2003) ‘Extreme Weather Events likely to Cause Disruption to Electricity Distribution’,

Met Office Report prepared for the Network Resilience Working Group [20] Wareing B, (1999), ‘Evaluation of performance of AXCES and EXCEL cable at the

Shetland Test Site’, EA Technology Report No 5039.

[21] Wareing B, (2001), ‘Severe weather testing on AXCES cable at Deadwater Fell’, EA

Technologies Services Report T3550. [22] Wareing B, (2005), ‘Wood Pole Overhead Lines’, Institute of Electrical Engineers.


Project No. 82880

43

Appendix 1


Project No. 82880

44

Documents

Environmental Discretionary Reward Scheme SP Energy ... › userfiles › file › ...Environmental Discretionary Reward Scheme 2 SP Energy Networks 2015–2023 Business Plan Updated